Create our own data set with test cases from Leetcode:

• Each of us will select 10 random leetcode problems (4 easy, 4 medium, 2 hard)
• For each problem, manually alter the solution resulting in one logically incorrect solution and one syntactically incorrect solution
• Keep track of the problem number and the difficulty
• Link to Spreadsheet: https://docs.google.com/spreadsheets/d/1IVpkTNftBr5n0cqDMR8EBs-C95loa0s4usKulx7L6ug/edit#gid=0

Github

Solutions folder

• Each file is the base solution we used to modify the problem in a .txt file, the title is [problem number].txt

Altered_solutions folder

• Each file is the manually altered solution in a .txt, the title is [problem number]_s.txt or [problem number]_l.txt (s = syntactic change, l = logic change)

GPT_solutions folder

• Ran code from Altered_solutions on chat gpt. The GPT solutions from using our basic prompt is located in the basic_prompt folder. Each file has the naming convention [problem number]_[error type]_sol.txt